Using registry data to identify individual dairy cows with abnormal patterns in routinely recorded somatic cell...

Area of Science:

• Animal Science
• Veterinary Medicine
• Data Science

Background:

• Somatic cell counts (SCC) in milk are key indicators of intramammary inflammation and subclinical mastitis in dairy cows.
• Early detection of mastitis is crucial for animal welfare, herd health, and economic reasons, requiring efficient diagnostic tools.

Purpose of the Study:

• To develop and validate a method for differentiating normal from abnormal SCC patterns in dairy cows using routine registry data.
• To identify individual animals exhibiting SCC profiles suggestive of mastitis based on their historical data.

Main Methods:

• Utilized registry data from 13,996 Holstein cows across eight conventional herds from 2010-2020.
• Applied nonlinear regression models (Nonlinear Least Square, Nonlinear Mixed Effect) to fit log10-transformed SCC to days in milk (DIM) using Wood's curve function.
• Identified abnormal SCC patterns by detecting outliers in mean squared residuals (MSR) during model fitting at the animal level.

Main Results:

• Wood's style function provided a consistently better fit for log10-transformed SCC over days in milk compared to Wilmink's function.
• The MSR outlier detection method successfully identified individual animals with SCC curves deviating from their established normal patterns.
• This approach offers a potential tool for automated, registry data-based mastitis detection in large dairy herds.

Conclusions:

• A robust method was developed to identify abnormal SCC patterns indicative of mastitis in individual dairy cows using existing registry data.
• This registry data-driven approach holds promise for improving early mastitis detection and management in commercial dairy farming.
• Further implementation could enhance herd health monitoring and reduce the economic impact of mastitis.